Introduction
In August 2023, I sat sipping sweet tea with Tenzin, a Tibetan refugee, in her living room in a suburb of Vancouver.Footnote 1 Tenzin spoke of her two school-aged children, saying that when they are with their Tibetan friends, “I explain to them… the Tibetan language shouldn’t be vanished. Don’t speak English. You all come from India, and you know Tibetan”. Tenzin emphasized the children’s shared diasporic heritage in India, and referred to a single Tibetan language that, through the children’s decisions to speak English together, was at risk of disappearing. To which Tibetan language was Tenzin referring? Why did children’s uses of English threaten this language? Since 1959, when Communist China’s advancements onto the Tibetan plateau provoked widespread unrest, Tibetans have migrated across borders, bringing their languages with them. Linguists estimate that the Tibetan language family includes up to fifty-four grammatically distinct spoken languages that share a single literary language (Ward & Roche Reference Ward, Roche, Powers and Templeman2020:371; Tournadre & Suzuki Reference Tournadre and Suzuki2023:83–84). Much of Tibet’s language variation has been lost in diaspora. In addition to pressures to shift to dominant official languages including Hindi and English, speakers of diverse Tibetan mother tongues have adopted central Tibetan, a lingua franca based on Lhasa Tibetan. Amid this transformation to spoken languages, the literary language has endured as a resource for cultural survival. Diasporic Tibetans like Tenzin emphasize cultural unity by recognizing a single heritage language, ‘Tibetan’, represented by the literary language and incarnated in central Tibetan speech. This article examines how, throughout multiple migrations, young children draw on this objectified Tibetan heritage language to co-create metalinguistic awareness, or consciousness of language as a symbolic system. While Tenzin viewed her children’s use of English as evidence of language loss, I suggest that children orient to a Tibetan heritage language, even when they may not be speaking it.
Although the term metalinguistic awareness is often used to refer to the ability to segment, describe, and analyze language forms (Koda Reference Koda, Cook and Bassetti2005:311), it also entails a broader form of metalinguistic knowledge production—an objectification of language that is realized in the interactive uptake of linguistic features. As young children interactively negotiate meaning between codes, they rely on multimodal resources to associate cultural identities with language varieties. These multimodal interactive practices represent a form of metalinguistic knowledge that I term heritage language recognition, or an objectification of language that marks the shared culture of a minoritized language community. By examining everyday literacy activities in a Tibetan-Canadian family in metro Vancouver, I show that children displayed heritage language recognition through interactive patterns, including entextualization—or the isolation of language segments—and recontextualization—or the dialogic recycling and transformation of these language segments (Bauman & Briggs Reference Bauman and Briggs1990). Through these interactive strategies, children coordinated co-engagement through multimodal reference to literary objects such as books and writing utensils. Drawing from the methods of language socialization (Ochs & Schieffelin Reference Elinor, Bambi, Richard and Robert1984), I contextualize a detailed turn-by-turn analysis of conversational data from longitudinal video ethnography with interviews and participant observation to reveal that, despite mismatches in the language repertoires of adults and children, children use sound, touch, gesture, and objects to mediate a shared understanding of the Tibetan heritage language. Although anthropological theories of linguistic ideologies have asserted that metapragmatic discourse, or explicit rationalizations about language use, format the maintenance or transformation of linguistic structures (Silverstein Reference Silverstein, Clyne, Hanks and Carol1979:206–207; Mertz Reference Mertz2007:343), an exclusive focus on metapragmatic discourse as the mediator between linguistic form and social differentiation overlooks the role of young children in language change. Young children may not articulate explicit rationalizations of language functions, yet, through their participation in situated interactions, may actively shape the objectification and differentiation of codes. For example, Meek’s (Reference Meek2007:28–29) study found that children responded to adults’ code-switching between English and Kaska by associating Kaska and English with age-based identities. By analyzing heritage language recognition in interactive patterns of entextualization and recontextualization, this article demonstrates that young Tibetan children use their multimodal repertoires to display metalinguistic knowledge. I thus argue that heritage language recognition can emerge through children’s multimodal interactive strategies, even in the absence of elaborated metapragmatic discourse.
Tenzin’s directive to her children to speak in Tibetan rather than English acknowledges children’s metalinguistic knowledge more narrowly, in terms of code choice alone. This concern over code choice responds to the inter-generational loss of communicative competence in the recognized Tibetan heritage language. On the most recent Canadian census, 9,350 respondents named ‘Tibetan’ as their ethnic or cultural origin, while only 7,025 claimed ‘Tibetan’ as their mother tongue (Government of Canada, Statistics Canada 2022). Evidence suggests that such inter-generational language loss among heritage language communities is accelerating. While in the mid-twentieth century, immigrant communities in North America transitioned to monolingualism in the dominant language within three generations, in the twenty-first century, they are increasingly losing the heritage language within two generations (Au Reference Au, Brinton, Kagan and Bauckus2008:337). Although all forms of language shift complicate cultural continuity, language loss within the span of two generations places particular social pressures on children, who may struggle to achieve shared understanding with their parents due to mismatches in their code repertoires (Fillmore Reference Fillmore1991; Cho Reference Cho2000). By drawing on methods that center children’s multimodal engagement with the material world (Vygotsky Reference Vygotsky, Bruner, Jolly and Sylva1976; Mondada Reference Mondada2016; C. Goodwin Reference Goodwin2018), I demonstrate that, despite intergenerational language loss in Tibetan families, the heritage language continues to serve as an expressive resource for achieving shared understanding. Even when children rarely speak a Tibetan mother tongue and are still learning to read in any language, literacy activities support their metalinguistic awareness by providing a symbolic focal point for the elaboration of meaning in everyday conversation. This article suggests the need for multimodal approaches when examining the blurred boundaries between formal heritage language education and non-institutional practices with texts, thus advocating for a child-centered approach to the study of language ideologies that reveals forms of knowledge often overlooked by adults.
Methodology and the focal family
This study emerges from my long-term inquiries into children’s agentive roles in reformatting Tibetan linguistic diversity in historically particular situations of migration across the Tibetan plateau. Such inquiries require careful attention to language use in embodied interaction, because young children’s communicative repertoires remain grounded in sounds, gestures, and engagement with material artifacts that often defy grammatical regularities, thus challenging the examination of diachronic linguistic change within children’s speech production alone. The linguistic anthropological study of literacy as a dialogic activity that is framed through sociohistorical and ideological norms has been particularly important in collapsing the boundaries between speech, text, and embodied interaction (Gutiérrez Reference Gutiérrez2008; Gee Reference Gee2014; Debenport & Webster Reference Erin and Anthony2019). As Gutiérrez, Baquedano‐López, & Tejeda (Reference Gutiérrez, Baquedano‐López and Tejeda1999:287) suggest, moments of social tension in literacy activities create hybrid spaces in which ‘the object of activity is extended and the activity itself reorganized, resulting in new opportunities for learning’. The examination of tensions within literacy activities, such as mismatches in code choice, reveals how participants co-create new understandings of language as the focal object of an activity—understandings that inherently involve multimodal resources reaching beyond speech or text alone. By centering children’s multimodal engagement with entextualized language, this study illuminates how a Tibetan heritage language is being constructed through blurred boundaries between text and talk, in a new setting of language contact between Tibetan and Canadian English. The refugee families who participated in this study had been living in protracted exile in India prior to migrating to Canada; they moved between a setting where Tibetan is formally taught in government schools to one that relies on community-based weekend schools. In this ethnographic setting, I offer the concept of heritage language recognition in order to analyze how young children objectify and differentiate codes through their multimodal communicative repertoires, and thus identify as speakers or non-speakers of their heritage language, without offering metapragmatic rationalization.
In order to analyze heritage language recognition in this setting, I contextualize ‘interaction analysis’, or the examination of sequences of embodied talk, based on twelve months of video ethnography with one focal family, alongside excerpts from ethnographic interviews and participant observation in a Tibetan heritage language school. Drawing on foundational principles from conversation analysis, interaction analysis examines the turn-by-turn unfolding of face-to-face talk alongside participants’ embodied movement and broader engagement with the material environment (M. H. Goodwin Reference Goodwin1991; Jordan & Henderson Reference Jordan and Henderson1995:40; C. Goodwin Reference Goodwin2018). While interaction analysis focuses on the sequential constitution of shared understandings in real-time talk, the social actions revealed through interaction analysis rely on accumulated histories that shape frameworks of taken-for-granted knowledge, including metalinguistic knowledge and language ideologies. Scholarship from linguistic anthropology emphasizes that language ideologies objectify language as a resource for meaning-making, shape cultural routines that make salient certain linguistic features, and drive social and linguistic differentiation (Kroskrity Reference Kroskrity, Schieffelin, Woolard and Kroskrity1998; Woolard Reference Woolard, Schieffelin, Woolard and Kroskrity1998; Makihara & Schieffelin Reference Miki, Bambi, Makihara and Bambi2007; Hauck Reference Hauck2018). However, despite the instantiation of language ideologies in interaction, and the interest of interaction analysis in accumulated histories of shared knowledge, these two bodies of scholarship have tended to remain separate in linguistic anthropological analyses. The study of metalinguistic knowledge in language ideologies has traditionally relied on an analysis of explicit metapragmatic discourse, not directly intersecting with the study of embodied sequential interaction.
To centre how children’s metalinguistic awareness becomes apparent through sequential interactive moves, rather than explicit rationalizations of the social value of codes or language features, I closely examine the language use of children in one family and community. This community includes approximately two hundred Tibetan refugees from Arunachal Pradesh, India, who resettled in metro Vancouver between 2014 and 2019 through a federal refugee relief program. In 2015, the focal family, made up of Choekyi (mother), Jangchup (father), and then two-year-old daughter, Tenzin, arrived in New Westminster, British Columbia. In 2018, the focal child, Khandro, was born. From March 2021 to March 2022, the family created weekly video recordings approximately one-hour in length, documenting spontaneous language use during everyday family activities in their home. Khandro’s language use, from ages two years eleven months to three years eleven months, formed the primary data for interaction analysis. In total, the video corpus included sixty-four videos created over fifty-two weeks, and featured activities including eating, playing with toys, reading and writing, drawing, and singing. The family’s independent creation of the videos allowed the research to continue in the context of the Covid-19 pandemic. Each week, one parent met with me on zoom to transcribe and analyze the recordings. Using ELAN, Tibetan talk was transcribed through broad (phonemic) transcription in the International Phonetic Alphabet (IPA), and English talk was transcribed with English orthography. Meaningful sound irregularities from child speakers in both Tibetan and English were transcribed narrowly in the IPA. Paralinguistic details of speech and turn-taking behaviours were documented using the conversation analysis conventions established by Gail Jefferson.Footnote 2 During transcription, I noted the parents’ reflections on their daughters’ language use, which became an important part of the ethnographic record. I conducted additional detailed interaction analysis of the sequencing of embodied conversational moves beyond these transcription sessions. Following Mondada (Reference Mondada2016), I incorporated video stills into transcript segments, indicating their simultaneous sequencing with language forms. In August 2023, I conducted nine ethnographic interviews with adult members of the community. The interviews were conducted in Tibetan and English, transcribed and translated using ELAN, and qualitatively analyzed through thematic discourse analysis with NVivo. In August 2023, I also attended the community’s weekly Tibetan heritage language classes and took detailed fieldnotes during participant observation.
In heritage language classes and the focal family home, reading and writing served as the primary activities in which adults explicitly stated what children shes ‘know’ about language, thus demonstrating the belief that literacy represents authentic metalinguistic knowledge. It was also through entextualized language forms, such as reading, writing, and singing, that children displayed metalinguistic knowledge in heritage language classrooms, and that Khandro and Tenzin most frequently produced Tibetan language forms in their home. In the corpus of video ethnography, Choekyi and Jangchup spoke in central Tibetan, with code mixing from English. Khandro and Tenzin, in contrast, spoke almost exclusively in English. While the children sometimes used central Tibetan to read, write, and sing, they always used English to frame and describe the Tibetan heritage language. Beyond activities explicitly organized around literacy, the children used central Tibetan in prompting routines, often drawing on books and other literacy materials that happened to surround the focal speech event. Literacy activities and literacy materials therefore offered a lens for examining children’s production of the Tibetan heritage language, as well as the metalinguistic awareness that they displayed through embodied conversational moves.
While literacy activities presented a primary context for children’s production of Tibetan, they were also salient in the corpus more generally. Of sixty-four total video samples, thirty-one centrally involved the canonical literacy activities of reading and writing. Choekyi and Jangchup chose when and what to record, with the stated goal of helping the researcher examine Khandro’s language learning. Their choice to record literacy activities may have been shaped by cultural norms that govern language use with young children in this community more broadly. For example, in heritage language classes, children wrote and rewrote themed words, such as the names of colours, that were linked to cultural artifacts, such as prayer flags, and elaborated in songs and activities. At home, parents oriented to reading and writing in both English and Tibetan, in order to supplement the offerings of Canadian schools. In interviews, parents mentioned that Canadian schools did not provide children with homework, which, in Tenzin’s words, meant that “there is no way kids learn”. Instead, she and her husband spent one to two hours daily instructing their children in “Tibetan handwriting and Tibetan language”, linking writing to recognizable metalinguistic knowledge that would support her children’s education more generally. An orientation towards formal education meant that parents sought out opportunities for their children to practice reading and writing, and engaged in extensive prompting routines that scaffolded children’s formulaic displays of knowledge of the literary language.
Reading throughout resettlement: Literacy and language ideologies among Tibetans from Arunachal Pradesh in British Columbia
The history of the Tibetan diaspora informs Tibetan-Canadian parents’ interest in developing their children’s literacy skills in both English and Tibetan. In the 1950s, Communist China’s westward advancement, and the resulting encroachment on Tibet’s existing political system of monastic governance, sparked social turmoil. In March 1959, Chinese military interventions in Lhasa, Tibet’s capital city, led Tibetans’ spiritual and political leader, the fourteenth Dalai Lama, to seek refuge in India. An estimated 80,000 Tibetans followed their leader into exile. With extensive international aid, the Dalai Lama established a sovereign government-in-exile based in Northern India, known as the Central Tibetan Administration (CTA). The CTA oversees a series of fifty-eight settlements located across South Asia (Brox Reference Brox2016:3), as well as a network of schools that provide Tibetan- and English-medium education (Bangsbo Reference Bangsbo, Barnett and Schwartz2008). Since the late 1960s, the United Nations High Commissioner for Refugees (UNHCR) and the CTA have negotiated for the resettlement of exiled Tibetans to Canada, the United States, Australia, and Europe. Tibetans living in South Asian settlements lack citizenship, face economic hardship, and experience ethnic discrimination. Tibetan settlements in the northeastern state of Arunachal Pradesh are particularly marginalized. In this contested region, China’s and India’s competing claims to territorial sovereignty exacerbate internal conflicts over land rights that often result in calls for the expulsion of Tibetan refugees (Tibet Justice Center 2016).
Given the vulnerable status of Tibetan refugees in Arunachal Pradesh, the UNHCR and CTA lobbied the Canadian government to accept Tibetans from this region on humanitarian grounds. In 2011, the Canadian federal government agreed to resettle 1,000 Tibetans from Arunachal Pradesh (Immigration, Refugees and Citizenship Canada 2011), most of whom had existing knowledge of English through education in Tibetan exile schools. Approximately two-hundred Tibetans from the resettlement program were sent to metro Vancouver, primarily the suburbs of New Westminster and Burnaby. While spots in the program were highly coveted, choosing to participate meant leaving behind aging family members, taking on the financial responsibilities of moving to a city with an extremely high cost of living, and raising children in a largely non-Tibetan social environment. These challenges are shared among many inhabitants of metro Vancouver. Tibetans from Arunachal Pradesh joined a metropolitan area where 41.8% of inhabitants were born outside of Canada, and 42.3% identify a mother tongue other than the official languages of English or French (Government of Canada, Statistics Canada 2022). Given the representation of new immigrants in the province, the education system in British Columbia aims to support English language learners, largely through formal instruction in reading and writing.
In fact, in ethnographic interviews, Tibetan parents from Arunachal Pradesh stated that access to free public education served as one of their primary motivations for migration. For example, Wangmo, the mother of a six-year-old boy, mentioned her gratitude for being able to access schooling as a permanent resident, when many international students pay expensive tuition fees to attend Canadian high schools and universities. Given their interest in formal education, Tibetan parents sought out additional forms of early English literacy instruction for their children. In metro Vancouver, private preschools and daycares were often cost prohibitive, but parents participated in StrongStart, a free program for all parents of children five and under operated through the province of British Columbia.Footnote 3 Along with one caregiver, young children attended StrongStart programs, where they engaged in group play activities including English letter recognition, drawing, singing, and arts and crafts. Tibetan families also participated in the non-profit-run Home Instruction for Parents of Preschoolers, where a trained Tibetan volunteer provided mothers with a curriculum of storybooks and activity packs to use with their children for fifteen to twenty minutes per day, in order to prepare them for entering kindergarten.Footnote 4 These programs introduced Tibetan parents to the norms of learning promoted in British Columbia’s public schools, and provided them with models for literacy activities at home. As Yangmo, the mother of a seven-year-old daughter explained, at StrongStart, “it was nice seeing other people, how they’re doing… like how the lives are here. It’s so different [from India]”. These Canadian programs thus provided pathways for more general social integration, while also defining learning primarily through literacy activities.
Although education represented a key factor driving parents’ decisions to migrate to Canada, parents also expressed concern that life in Canada would dissociate their children from their linguistic and cultural heritage. In fact, in ethnographic interviews, many adults reported that Tibetan-Canadian children speak Tibetan rarely, and all reported that their children prefer to speak in English. Some children born in Canada initially spoke in Tibetan before shifting to English. For example, Dolma, the mother of two young adult children, explained “After a couple of years, a year and a half, two years, they started talking, totally speaking Tibetan….When I sent them to preschool, then they started learning English, and they always speak English”. While for many children, relationships with English-speaking peers led to a shift in their language repertoires from Tibetan to English, other children, like Khandro, primarily used English from the time they began to speak.
In 2017, the community developed the weekly Tibetan heritage language school run by volunteer teachers in order to intervene in children’s seeming loss of Tibetan. The volunteer teachers focus their instruction on literary Tibetan, building lessons from writing prompts oriented around a theme or cultural activity such as Buddhist prayers, animals, or food. The goal of using literacy to maintain intergenerational language transmission is complicated by the fact that, in a situation of classic diglossia, mutually unintelligible spoken varieties are united by a single literary language. Tibetans from Arunachal Pradesh living in metro Vancouver speak central Tibetan, as well as English, Hindi, and Nepali. Some community members speak additional Tibeto-Burman languages such as Pemako (Tshangla), Tawang Monpa (Takpa), and Sherpa, based on their particular family backgrounds and migration histories. The community’s focus on developing literacy skills for the purposes of language survival seems to overlook the diverse spoken language varieties that Tibetan-Canadian children are exposed to in the home, as well as the non-standard communicative repertoires of emerging child speakers.
Despite variation within Tibetan, parents and volunteer teachers defined communicative competence through knowledge of the Tibetan literary language. For example, Paldon, a volunteer teacher at the Tibetan heritage language school in Vancouver, suggested that the literary language supersedes the diversity of spoken languages. When describing the importance of literacy instruction for her students, Paldon explained that, if the children learned to read and write, their Tibetan speech would be gsal po ‘clear’. With the adjective ‘clear’, Paldon suggested that the literary language represents a model variety that is comprehensible to Tibetans from different regions. Tibetan scholars have similarly looked to the classical literary language not simply as an ideal standard, but as a map to a comprehensive understanding of Tibet’s diverse spoken languages. When commenting on features of spoken linguistic diversity that are manifested in literary Tibetan, for example, scholar Sangye Tandar Naga (Reference Naga2009:561) asserted, ‘if we are able to discover the mystery prevailing in different dialects of the Tibetan language, everyone who knows written Tibetan [could] speak all different Tibetan dialects’. While literary Tibetan is framed as a unifying standard language, in the Tibetan diaspora, central Tibetan remains the most frequently used spoken language and serves as the medium of instruction in heritage language schools. Although the Tibetan heritage language is represented in central Tibetan speech, adults discussed reading and writing as the primary form of metalinguistic knowledge—that is, as demonstrating a comprehensive understanding of the Tibetan language that transcends any particular spoken variety.
The Tibetan community’s focus on the literary language as an embodiment of a pan-Tibetan language demonstrates the centrality of relationships between text and talk in formatting metalinguistic knowledge. Studies of reading and writing in heritage language classrooms in other ethnographic settings have analyzed the construction of authority through linguistic resources such as repairs (He Reference Agnes Weiyun2004) and markers of politeness (Byon Reference Byon2003) that rely on the circulation of entextualized language forms. These studies of entextualized language demonstrate that, in the course of dialogically producing metalinguistic knowledge, participants isolate language segments, assign meanings to linguistic features based on code distinctions, and differentiate language from other communicative resources. Although scholars recognize that participants’ communicative repertoires integrate resources from across codes as well as across modalities, participants actively build language boundaries in the course of producing metalinguistic knowledge about heritage languages. Specifically, informed by ideologies that defined metalinguistic knowledge through literacy, Tibetan-Canadian adults marked English and Tibetan as distinct languages, and constrained children’s uses of Tibetan to second-pair responses to questions in prompting routines. Children, however, creatively blurred this boundary between English and Tibetan, using material artifacts to draw equivalencies between English and Tibetan sounds, phrases, words, and letters.
While writing systems serve as a foundation for articulating and producing language ideologies in a range of ethnographic contexts (Jaffe, Androutsopoulos, Sebba, & Johnson Reference Jaffe, Androutsopoulos, Sebba and Johnson2012), beliefs about literacy can become particularly elaborated in diasporic communities like Tibet-in-exile, where continued senses of displacement and hopes for sovereignty in the homeland intersect to shape the formation of belonging and difference. As Sonia Das (Reference Das2016:163–64) found in an ethnography of Sri Lankan and Indian Tamil-speaking refugees in Montreal, the use of the literary language in heritage language education formatted distinct relationships to the homelands of India and Sri Lanka that drew on political ideologies of linguistic difference and reproduced a boundary between Sri Lankan and Indian Tamils in their new home in Montreal. Das’ work suggests that, through language ideologies oriented towards texts, language standardization and linguistic differentiation inform one another. In other ethnographic settings, scholars have noted that standard language ideologies, including an orientation towards a literary language, can unintentionally hinder the full participation of heritage language learners (Wong & Xiao Reference Wong and Xiao2010; De Korne Reference De Korne2017). These studies suggest that overt support for linguistic diversity enhances revitalization. However, while the valorization of linguistic unity may complicate Tibetan-Canadian children’s development of spoken communicative competence, the literary language also provided material resources for children to develop their identities as heritage language speakers. Conversational activities always recall an accumulation of historical associations, which evolve through judgments about the value of certain forms of knowledge and serve to define the possible roles of persons as well as language varieties (Irvine Reference Irvine, Silverstein and Urban1996; C. Goodwin Reference Goodwin2018). Through the multimodal recontextualization of language forms in everyday talk, young Tibetan-Canadian children revealed their judgments about who could claim to speak Tibetan, as well as how they could speak Tibetan.
Interaction analysis
By employing interaction analysis of Khandro’s everyday family conversations, this article asserts two claims. First, literacy activities provide material anchors for Tibetan children to build their identities as heritage language speakers. Second, we can analyze children’s metalinguistic knowledge through their multimodal engagement with entextualized language forms. In the course of literacy activities, Jangchup and Choekyi offered entextualized language segments as models for Tenzin and Khandro’s language production. While Jangchup and Choekyi used prompting questions as a broader conversational resource, in literacy activities, they expected Tenzin and Khandro to respond in Tibetan rather than English. The children, however, used English to comment on Tibetan language forms, and used books and toys to elaborate on the meanings of segments of spoken language. That is, Tenzin and Khandro recontextualized isolated language forms, expressing alignment with Jangchup and Choekyi’s interactive goal of supporting their Tibetan language use. In the following excerpts, the children use tools from the material environment to assist in their production of Tibetan language forms.
Excerpt (1): “He’s saying it”
In late March 2021, Khandro was two years, eleven months old. Khandro was sitting with Jangchup in the family bedroom. In an inversion of their typical parent-child roles, Khandro had initiated a question-answer sequence, asking Jangchup to name colours on her clothing. After twelve question-answer pairs, Jangchup initiated a new activity of singing. Jangchup said, ŋa-t͡sʰo ʃe-taŋ-kʲa a? ʃe-taŋ-ki-jin-pe taŋ-ki-min ‘We’ll sing a song, eh? Will you sing or not?’. Jangchup then switched into English in the same utterance, “You wanna sing a song?”. Khandro responded by calling out happily, “Yeah!”.
Jangchup began to utter single verses of a song, entextualizing each as a discrete language segment. Khandro repeated after him. Khandro jumped off the bed and retrieved a stuffed dog named Snowflake. As she brought the toy onto the bed, she also brought it into the frame of her interaction by voicing the stuffed dog’s sounds. One minute and ten seconds after Khandro brought the stuffed dog onto the bed, Jangchup reoriented the toy’s contributions, through Khandro, towards the singing activity.
(1) “He’s saying it”; J: Jangchup, K: Khandro
Jangchup instructs dog.
Khandro voices dog.
Khandro voices dog.
Jangchup instructs dog.
Dog faces Jangchup.
Khandro completes verse.
In lines 1–4, Jangchup drew on Khandro’s interest in the stuffed dog, Snowflake, to maintain alignment in the interactive goal of singing a Tibetan song. Jangchup first used English to direct Snowflake to sing (line 1). Khandro spoke, in her own voice, in English, accepting Jangchup’s proposition (line 2). Jangchup then provided additional, multimodal directives. Using Tibetan, he instructed Khandro and Snowflake to “read” with him, and look at him (line 3). Jangchup’s framing of the act of singing as “reading” highlights the literary register of the Tibetan song. It also emphasizes the interactive structure of prompting and repetition, which occurs in the formal instruction of reading and singing. In response, Khandro shifted to a high pitch, showing that she was voicing Snowflake, and indicated agreement with Jangchup’s directive (line 4).
In lines 5–6, Jangchup, Khandro, and Snowflake collaboratively repeated a verse of the song. In line 5, Jangchup provided the first words of the verse as a prompt, indicated with a rising tone. He pointed to Snowflake as he spoke, suggesting the need for the dog’s active participation (Figure 1a). In line 6, Khandro completed the verse by providing the correct Tibetan verb form. She shifted into a high pitch and moved the stuffed dog’s head, animating Snowflake’s voice (Figure 1b).
After one minute and fifty seconds, Jangchup reoriented the singing activity. Instead of providing initial prompts, Jangchup used English to suggest that Khandro teach the song to Snowflake (line 7). Khandro used format tying, or the repetition of a prior segment with transformation (M. H. Goodwin Reference Goodwin1991:180), to provide a preferred second to Jangchup’s suggestion. She began the verse in her own voice, and then brought her face close to Snowflake’s and shifted her pitch to animate the stuffed dog’s voice (line 8, Figure 1c). That is, Khandro provided the beginning of the verse as a prompt in her own voice, and then completed the verse in Snowflake’s voice. In line 10, Khandro provided a metalinguistic gloss of her prior turn, explaining in English that Snowflake was “saying” the song.
In lines 11–18, Jangchup advanced the singing activity through entextualization and recontextualization. Jangchup provided two-syllable segments of the verse as models for repetition. In line 11, he pointed to Snowflake, marking the dog as his conversational partner (Figure 1d). In line 12, Khandro shifted into a high pitch and moved Snowflake’s head as she repeated the syllables, showing that she was enacting the dog’s voice. In line 13, Jangchup mirrored Khandro’s prosody by shifting into a high pitch, enacting the expected response in the dog’s voice. Khandro retained this high pitch in line 14. In line 15, Jangchup provided the final two syllables of the verse. In line 16, Khandro repeated these syllables while turning to face Jangchup, and then moving slightly away from him. In line 17, Jangchup advanced the activity sequence by directing Khandro and Snowflake to sing the complete song. He then repeated the full verse. As he sang, Khandro moved Snowflake to face Jangchup and shook Snowflake’s head, maintaining the dog’s active participation (Figure 1e). In line 18, Khandro joined Jangchup in reciting the final two syllables of the verse (Figure 1f).
As Mondada (Reference Mondada2016:341) asserted, while any material details of an interactive setting can serve as resources for the elaboration of action, some resources may be more conventionalized than others. Khandro and Jangchup used entextualized language forms in a conventional act of singing, which they glossed as reading through metalanguage as well as turn-by-turn prompting. This conventionalization served to define the heritage language, embodying it in the words of the song. Despite the formulaic nature of this entextualization, Khandro demonstrated her agency in claiming knowledge of the heritage language by creatively appropriating other resources as she recontextualized language segments. Vygotsky (Reference Vygotsky, Bruner, Jolly and Sylva1976:546) wrote of play activities as opportunities for the acquisition of new skills, which arise because children attribute novel and unexpected meanings to objects.Footnote 5 In this excerpt, Khandro used an object to playfully articulate metalinguistic knowledge. She drew on a stuffed dog, along with movement, pitch, and gesture, to imaginatively build a new speaker through Snowflake’s voice. She also positioned herself as an expert, providing initial prompts as the basis for Snowflake’s repetition. Through multimodal elaboration in play, Khandro therefore acted as a heritage language speaker, beyond providing second-pair parts to conventionalized prompting questions. Ten months later, Khandro again defined herself as a heritage language speaker, in this case through her use of English.
Excerpt (2): “‘Cause I’m one and I cannot read”
In mid-January 2022, Khandro was three years and nine months old. She was eating a meal with sister Tenzin, while Choekyi prepared food in the kitchen. After eating for about ten minutes, Khandro stood up and announced, “I want to draw!”. With Choekyi’s help, she brought a colouring book onto the kitchen table (Figure 2a).
Khandro retrieves colouring book.
As Tenzin continued to eat, Choekyi said the first letter of the Tibetan alphabet, and instructed Khandro to repeat after her. Khandro protested that she was drawing, but Tenzin encouraged her to join Choekyi’s activity. Tenzin called, “but you can also do the practice of the other ones too, right? Come on, keep her going, Mom!” (Figure 2b).
Tenzin directs Choekyi to initiate prompting sequence.
In response to Tenzin’s call, Choekyi initiated a prompting routine of saying the alphabet. She named each letter, and the girls repeated after her. Midway through the alphabet, Choekyi switched the form of activity, segmenting sentences and prompting Khandro and Tenzin to repeat words and phrases back to her in central Tibetan.
(2) “‘Cause I’m one and I cannot read”; C: Choekyi, T: Tenzin, K: Khandro
Tenzin counts age silently.
Tenzin begins to count Khandro’s age out loud.
Tenzin counts Khandro’s age.
Khandro explains her age.
In this conversational sequence, Choekyi entextualized Tibetan language forms through prompting questions. Khandro’s action of retrieving the drawing book seems to have provoked the start of the activity sequence; it was after Khandro settled at the table with her drawing book that Choekyi began to recite the alphabet. Although Khandro initially resisted, Tenzin explained that Khandro could “practice” speaking and drawing at the same time, and enthusiastically called on Choekyi to keep up with her questioning (“keep her going, Mom!”). The activity then moved to a series of prompting questions, in which participants’ layering of multimodal resources became particularly elaborated.
In lines 1–4, Choekyi offered single Tibetan words that Tenzin repeated back to her. Khandro continued to draw. In line 5, Choekyi provided a full sentence as a question, and repeated it. In line 6, she used English to clarify the meaning of the Tibetan word year and then repeated the question in English. Meanwhile, Tenzin was silently counting on her fingers. In line 7, she concluded that she was eight years old. In line 8, Choekyi acknowledged Tenzin’s correct answer, and then directed the question towards Khandro. Tenzin began to use Tibetan to count out loud on her fingers (line 9), but Khandro interrupted her to announce “four” (line 10). In lines 11–13, Tenzin and Choekyi explained to Khandro her age. Tenzin told Khandro that she is not four (line 11), but then Choekyi directed Khandro to produce the Tibetan word for four (line 12). In line 13, Tenzin counted to three in Tibetan, holding up one finger for each number.
Then, the girls veered away from the activity sequence and began to gently wrestle with each other. One minute and twenty-seconds later, Choekyi restarted her prompting questions, shifting the girls’ attention (line 14). In line 15, Khandro responded with the word “good”. Tenzin and Choekyi then overlapped one another to correct her. Tenzin called out “three” in Tibetan (line 16), while Choekyi repeated her question in English, and then answered in Tibetan (line 17). In line 18, Tenzin expanded on her contribution from line 16, explaining to Khandro that she is three years old. Khandro, however, did not accept Choekyi and Tenzin’s suggested answers. In line 19, Khandro momentarily paused her drawing activity, looking up and explaining that she is one year old because she cannot read.
As Mondada (Reference Mondada2013:606) observed in a study of guided visits, ‘displays of knowledge and epistemic authority are achieved within a specific action format and sequence organization’, involving conventionalized initiations of topical sequences. In Tibetan-Canadian family routines, books figured centrally as devices for initiating question-answer sequences, with knowledgeable participants crafting questions about letters, words, and images. In this sequence, the participants drew on Khandro’s activity of drawing to initiate prompting questions that focused on entextualized Tibetan language forms. That is, Khandro’s initial action of retrieving a book seems to have provoked the prompting routine. Even though Khandro’s action may have begun the sequence, Choekyi and Tenzin used correction to portray Khandro as lacking in epistemic authority. Choekyi moved between English and Tibetan, explaining to Khandro that she should answer in Tibetan (line 12), and Tenzin used correction as well as gesture to amend Khandro’s answer by providing her true chronological age (lines 11, 13, 16, 18). Khandro, however, displayed her own epistemic authority by resisting these corrections. Khandro claimed authoritative knowledge that she is one year old by describing her inability to read (line 19). Khandro therefore ironically drew on ideologies that associate literacy with knowledge (and illiteracy with a lack of knowledge), in order to assert her ability to independently answer Choekyi’s prompts. Khandro’s activity of drawing, as well as her overt claims to a lack of knowledge, allowed her to display authority in the collaborative construction of prompting questions.
As they recontextualized language forms in prompting routines, Tenzin and Khandro displayed heritage language recognition by using multimodal resources to act as speakers of Tibetan. The girls also actively located equivalent language forms in Tibetan and English. Attention to the linguistic form of conversational turns would suggest that children were using the dominant language, English, to analyze the heritage language, Tibetan. The examination of multimodal conversational turns, however, suggests a more holistic mediation of metalinguistic knowledge. In the sequential unfolding of conversations, the girls expanded on prior turns to build equivalencies that relied on material artifacts in books and drawing materials. As Vygotsky (Reference Vygotsky, Cole, Jolm-Steiner, Scribner and Souberman1978) suggested, children’s metalinguistic knowledge emerges in situated activity contexts that are inherently social and material. That is, as children interpret shared meanings, ‘the path from object to child and from child to object passes through another person’ (Reference Vygotsky, Cole, Jolm-Steiner, Scribner and Souberman1978:30). Metalinguistic knowledge is therefore first instantiated in children’s social interactions with the material world. Prompted by the introduction of physical literacy materials, Tenzin and Khandro then engaged in collaborative expansion involving the layering of sonic, visual, haptic, and linguistic resources. Through multimodal resources, the children creatively negotiated metalinguistic knowledge and defined themselves as speakers of the Tibetan heritage language.
Excerpt (3): “Which one is what in English”
In late March 2021, Jangchup was sitting in bed with two-year-eleven-month-old Khandro, reading an English children’s book. Soon, Tenzin entered the bedroom and joined Jangchup and Khandro on the bed. One minute and twenty seconds later, Jangchup announced, in Tibetan, da ŋa-t͡sʰo pɶ-jik tis d͡ʒɑŋ-go lo ‘Let’s study a little Tibetan, ok?’. Jangchup reached for an alphabet book and a pencil on a nearby table. Jangchup nestled Khandro in his lap, placing the book in front of them while Tenzin viewed the open book from his right-hand side. With this ‘corporeal configuration’ (de León Reference de León, Duranti, Ochs and Bambi2014:85), Jangchup created an affective and sensory alignment with Khandro, which allowed Khandro to take up a similar stance of authority towards the act of reading. Jangchup held the pencil, and pointed to each letter in the Tibetan alphabet as he said it out loud. The girls repeated the letters after him. Four minutes and two seconds later, they reached the end of the alphabet. Tenzin began to draw equivalencies between English and Tibetan letters, and Khandro claimed active knowledge through her manipulation of the pencil and book.
(3) “Which one is what in English”; J: Jangchup, T: Tenzin, K: Khandro
Tenzin points and names letters.
Khandro points with Tenzin.
Tenzin notes the beginning and end of the alphabet.
Khandro grabs pencil.
Khandro prepares to instruct.
Tenzin explains.
In line 1, Jangchup concluded the activity sequence of reading the alphabet with a positive assessment of the girls’ performance, first in English and then in Tibetan. In line 2, Tenzin established mutual orientation towards the equivalency between English and Tibetan, by reaching and pointing to the final two letters of the Tibetan alphabet and naming them with the English letters, Y and Z (Figure 3a). Khandro took up Tenzin’s action, repeating “Y Z” and using the pencil to point to the final two Tibetan letters (Figure 3b). In line 4, Tenzin began an account of her use of English letters, again using body alignment to achieve shared attention as she moved her hand to point to the first two letters (Figure 3c). Jangchup responded with an affirmation (line 5). Tenzin latched on to continue her account, explaining the equivalency between the first letters of the alphabets (line 6). While the other participants did not acknowledge Tenzin’s account, Khandro modified the participants’ body alignment to advance the activity sequence. She grabbed the pencil, which Jangchup had tried to take away from her, and pushed away Jangchup’s arm. She overlapped Tenzin, asserting that it was “her turn” (line 7, Figure 3d). Over the next minute and nineteen seconds, Khandro and Tenzin repeated the name of each Tibetan letter after Jangchup. Khandro continued to assert her knowledge by moving the pencil to point to each letter in turn.
In lines 8–17, both girls actively displayed their knowledge of Tibetan through English equivalencies. In line 8, Tenzin restarted her action of associating English letters with Tibetan letters based on their order, beginning in the middle of the alphabet. Jangchup overlapped Tenzin, asserting that her naming of the letters was not correct (line 9). Khandro overlapped Jangchup and Tenzin’s dyad, using the pencil in her left hand, along with her right finger, to point to the letters. Khandro then instructed the others, positioning herself both as the authority (“telling”) and recipient (“tell me”) of knowledge of the alphabet (line 10, Figure 3e). Tenzin overlapped Khandro with an interjection, protesting Jangchup’s prior assertion (line 11). In line 12, Jangchup latched onto Tenzin’s interjection, elaborating on his assertion. Tenzin acknowledged his explanation (line 13), and Jangchup expanded his explanation, stating that ABCD are English letters (line 15). Tenzin overlapped him, providing a final account of her action by stating that she was explaining to Khandro the equivalency between English and Tibetan letters (line 16, Figure 3f). Meanwhile, Khandro overlapped Tenzin and Jangchup, taking on the authoritative role of initiating letter repetition by pointing to the first letter and calling out “A” (lines 14, 17).
In this sequence, Tenzin and Khandro did not simply repeat equivalent content in Tibetan and English, but used visual, gestural, and haptic resources to bridge their knowledge of the two languages. Morgenstern (Reference Morgenstern and Bolens2023:27) writes of children’s language acquisition that ‘the sense of agency that [children] experience through their bodies in motion in the mundane activities of their daily lives resonates with their environment filled with people, dynamic events, and artifacts that they perceive and organize with their burgeoning sensorimotor skills’. That is, children rely on touch and movement to agentively coordinate shared attention. In this excerpt, Tenzin and Khandro relied on the images of letters, as well as pointing gestures, to claim their actions as legitimate displays of metalinguistic knowledge. For example, Khandro insisted on using the pencil to point to letters throughout the entire sequence, pushing away Jangchup’s hand and taking an active role in identifying each letter rather than simply repeating it. She also used Tenzin’s creation of letter equivalencies to initiate repetitions, pointing to the first letter on the page and repeatedly naming it as “A”. Pointing functioned as more than a referential device, allowing the speaker to explicitly exhibit their right to shape the next turns in the sequence (Mondada Reference Mondada2016:342).
These excerpts demonstrate that Tenzin and Khandro relied on simultaneous, multimodal resources to recognize the heritage language. Participants entextualized Tibetan language forms by segmenting verses, sentences, and letters. This entextualization revealed the ideology that the Tibetan language is manifested in a literary standard. Through recontextualization, participants claimed authoritative knowledge of how to analyze and use Tibetan language forms, positioning themselves as speakers of Tibetan even when they used English. While these claims to knowledge referenced the heritage language, itself, they also relied on material and haptic resources. For example, Jangchup and Choekyi drew on Khandro’s use of props, such as toys and colouring books, as material foundations to initiate prompting sequences. Both children also used touch, gesture, and material artifacts to actively recontextualize Tibetan language forms. Through the reflexive capacity of language, the children elaborated Tibetan language forms across multiple modalities, coordinating knowledge of the heritage language and its speakers that did not rely on metapragmatic discourse.
Discussion
Among diasporic Tibetan-Canadians, literacy serves as a foundation for metalinguistic knowledge. Ethnographic interviews as well as participant observation in a Tibetan heritage language school demonstrated that, amid efforts for cultural survival in diaspora, adults viewed literacy as a pathway to comprehensive understanding of the Tibetan language, as well as a means to intervene in intergenerational language loss. Despite participants’ explicit metapragmatic rationalizations regarding texts, their metalinguistic knowledge was mutually elaborated in the course of embodied interaction. That is, literacy ideologies afforded children with key modes of textual engagement that reached beyond formal reading and writing activities (Gutiérrez Reference Gutiérrez2008). Despite standard language ideologies that would seem to hinder children’s active use of diverse Tibetan language forms, multimodal resources provided children with the basis to objectify, analyze, and claim knowledge of Tibetan, even in the absence of fluent communicative competence. Children used sound, body position, and engagement with objects to entextualize and recontextualize language segments and thus actively recognize the heritage language.
Tibetan-Canadian children’s creative practices of entextualization and recontextualization demonstrate the need for multimodal approaches to examining the dynamics of heritage language maintenance and language shift. While previous studies have outlined the importance of literacy to heritage language maintenance, they have often focused on literate adults. For example, studying diverse immigrant communities in the United States, Kwon (Reference Kwon2017) described mothers’ strategies and motivations for heritage language maintenance, noting their everyday use of print literacy materials from their home countries. Also examining the ethnographic context of the United States, Tse (Reference Tse2001) found that young adults’ independent reading, as well as regular access to literary materials through community organizations, predicted their continued use of Spanish, Japanese, and Cantonese. Examining multilingual education in Canada, Duff (Reference Duff, Brinton, Kagan and Bauckus2008:76–77) noted that heritage language literacy, supported by parents, predicted better academic performance. While such analyses focus on the vitality of heritage languages, other studies have examined literacy following language loss. A heritage language learner of Tibetan, Tso (Reference Tso and Canagarajah2019:179), noted that methods of reading in English and Chinese enhanced her study of literary Tibetan. While these studies suggest that access to reading materials and the ability to read support heritage language vitality, in Tibetan-Canadian children’s everyday practices, texts provided affordances for language revitalization beyond conventionalized literacy.
Specifically, through the blurred boundaries between text and talk, children claimed identities as heritage language speakers, even when they lacked the ability to read and infrequently produced spoken Tibetan language forms. Instead, children used multimodal resources to recontextualize language segments, co-constructing the heritage language along with their parents. As Morgenstern (Reference Morgenstern and Bolens2023:23) outlines, the elaboration of knowledge in the course of language acquisition represents a form of ‘embodied cognition’, in which ‘understanding the meaning of words implies sensory, emotional, and motor experiences’. Children’s lived experiences of using language, rather than talk about language alone, co-constructed a culturally meaningful multilingual reality. Such a perspective on children’s creative recontextualization of heritage language forms allows for a shift in focus from the deficits of heritage language learners to the cultural practices of ‘new speakers’ who draw on diverse multimodal resources from their environments as they build new language practices (Soler & Zabrodskaja Reference Soler and Zabrodskaja2017:561). Analyzing multilingual speech through the lens of multimodal repertoires makes apparent possibilities for the revitalization of language repertoires, rather than static codes (Kusters, Spotti, Swanwick, & Tapio Reference Kusters, Spotti, Swanwick and Tapio2017). In fact, such a multimodal approach to language revitalization affords us with an expanded understanding of children’s metalinguistic knowledge that is particularly relevant to linguistically diverse diasporic communities. Children’s active construction and destabilization of language boundaries does not merely preserve a heritage language, but demonstrates possibilities for emplacing mobile linguistic and textual resources in new social and linguistic settings.
Attention to the multimodal co-construction of language within family literacy activities pushes the boundaries of what we believe children can know. Literature in language ideologies has traditionally approached metapragmatic discourse, or implicit and explicit commentary on language use (Woolard Reference Woolard, Schieffelin, Woolard and Kroskrity1998:4), as the interface between language form, function, and social identity. Tibetan-Canadian children’s interactive engagement with entextualized language, however, demonstrated the co-constitution of embodied metalinguistic knowledge that used multimodal resources to construct a valued heritage language. With a child-centered approach to language ideologies, we can identify forms of metalinguistic knowledge that are not readily recognized by adults, but nonetheless support the continued use of heritage languages, in new contexts, modalities, and functions, among migrant and diasporic communities.
Appendix: Transcription conventions
Interlinear glossing is simplified to show only Tibetan language forms and English translations.
Tibetan is bolded.
Children’s non-conventional sounds are underlined.
